omv-release-upgrade worked but issue with eth0 in the GUI

  • I did clone my omv6 system first at least and have reverted back so no drama

    I only hope all other users will take the same precaution as you did, ;)

    • Offizieller Beitrag

    Shouldn't this command be blocked until further tests were done? I mean, the ability to upgrade from 6 to 7?

    This will make a lot of users upgrade without considering the consequences of a BETA state OS.

    It actually doesn't cause a problem unless you delete the adapter from the web interface and try to fix it with omv-firstaid. My system's network connection is still working fine.

    omv 7.1.0-2 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.2 | k8s 7.1.0-3 | cputemp 7.0.1 | mergerfs 7.0.5 | scripts 7.0.7


    omv-extras.org plugins source code and issue tracker - github - changelogs


    Please try ctrl-shift-R and read this before posting a question.

    Please put your OMV system details in your signature.
    Please don't PM for support... Too many PMs!

    • Offizieller Beitrag

    So what is it?

    It looks like a normal adapter. Strangely, my other RPi with a fresh install of Raspberry Pi OS Bookworm is using eth0 not end0.


    2: end0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000

    link/ether dc:a6:32:f8:d6:03 brd ff:ff:ff:ff:ff:ff

    inet 192.168.1.190/24 metric 100 brd 192.168.1.255 scope global dynamic end0

    valid_lft 25608sec preferred_lft 25608sec

    inet6 fe80::dea6:32ff:fef8:d603/64 scope link

    valid_lft forever preferred_lft forever


    [ 8.251370] bcmgenet fd580000.ethernet end0: renamed from eth0

    [ 12.113956] bcmgenet fd580000.ethernet end0: Link is Down

    [ 16.192129] bcmgenet fd580000.ethernet end0: Link is Up - 1Gbps/Full - flow control off

    [ 16.192214] IPv6: ADDRCONF(NETDEV_CHANGE): end0: link becomes ready

    omv 7.1.0-2 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.2 | k8s 7.1.0-3 | cputemp 7.0.1 | mergerfs 7.0.5 | scripts 7.0.7


    omv-extras.org plugins source code and issue tracker - github - changelogs


    Please try ctrl-shift-R and read this before posting a question.

    Please put your OMV system details in your signature.
    Please don't PM for support... Too many PMs!

  • It did cause problems for me before I deleted anything in the gui.


    On my system end0 was down and eth0 was gone (ip a did not show it) but eth0 was partially displayed in the network page. Something was working as the rpi had an ip address.


    I deleted eth0 from the gui later. Then I lost the ip address / network

    • Offizieller Beitrag

    It did cause problems for me before I deleted anything in the gui.


    On my system end0 was down and eth0 was gone (ip a did not show it) but eth0 was partially displayed in the network page. Something was working as the rpi had an ip address.

    I just meant you (and I) could still access the web interface. eth0 shouldn't show in ip a since eth0 is renamed end0. eth0 does show on the network page because OMV is reading that from the database not from ip a. Your system still had an ip address because it seems netplan is "smart" enough to figure out that even though the netplan config is pointing to eth0, it knew that eth0 is now end0. So, people won't be locked out of their system unless they try to "fix" things.

    omv 7.1.0-2 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.2 | k8s 7.1.0-3 | cputemp 7.0.1 | mergerfs 7.0.5 | scripts 7.0.7


    omv-extras.org plugins source code and issue tracker - github - changelogs


    Please try ctrl-shift-R and read this before posting a question.

    Please put your OMV system details in your signature.
    Please don't PM for support... Too many PMs!

  • votdev

    Hat das Label gelöst hinzugefügt.
  • digging a bit more and I think OMV7 / Deb12 is creating a network interface called end0 rather than eth0.

    Re-reading this, when/if you retry the upgrade, can you check if you have disabled via raspi-config, the predictable names PRIOR to do the upgrade?


      



    I'll try to do this later today to test:

    Have a Pi4 running OMV6 and have it disabled.

    Run omv-release-upgrade and see how it goes.

  • 1st test; Pi4 4Gb RAM, OMV6 with 6.9.10-5 fully updated with only flashmemory & cputemp plugins.


    With Enabled predictable names via raspi-config.

    ip a shows eth0:

    Code
    pi@sinkhole:~ $ ip a
    ---------
    2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
        link/ether dc:a6:32:54:5e:fd brd ff:ff:ff:ff:ff:ff
        inet 192.168.1.191/24 brd 192.168.1.255 scope global dynamic eth0
           valid_lft 3027sec preferred_lft 3027sec
        inet6 2001:818:e867:d900:dea6:32ff:fe54:5efd/64 scope global dynamic mngtmpaddr noprefixroute
           valid_lft 86384sec preferred_lft 43184sec
        inet6 fe80::dea6:32ff:fe54:5efd/64 scope link
           valid_lft forever preferred_lft forever


    Ran omv-release-upgrade inside a screen session

    After reboot, ip a still shows eth0: shows end0 [EDIT] Was still asleep, :sleeping: :sleeping: :sleeping: [/EDIT]


    Also, this time, the USER pi was added without error to the new GROUP _ssh.


    Upgrade went normal and without any issue.

    Only "fail" I have in this rig (nothing to do with OMV) is that an IC display stopped working.

    It's a Sunfounder PIRONMAN box and has some pip scripts to controller a fan, RGB leds and a display with some info.

    This is something that I need to check with the manufacturer to see if the scripts have been ported to Debian12.


    [EDIT]

    All it took was to reinstall the scripts with v2 from Sunfounder:
    5. Set up the Pironman — SunFounder Pironman documentation


    RGB strip, fan speed/temp and OLED display running again normal.

    [/EDIT]


    Next test will be with the Predictable names disabled.

  • What shows the network page in the GUI?

    I forgot to tell that eth0 is on DHCP.


    OMV7 GUI:


    omv-firstaid:


    Weird...

  • That is similar to what I see on the HC2.

    ip a shows a different name than the GUI and the netplan config.

    Rectified the info about ip a on #30. Didn't saw that it was showing end0 when already on OMV7.

    Funny thing is, I never lost connection to the Pi when doing the upgrade and the network was/is always ON and with the same IP.


    I'll flash back the OMV6 prior to the upgrade but this time will do it with the Predictable names disabled to see it it behaves the same.

  • I'll flash back the OMV6 prior to the upgrade but this time will do it with the Predictable names disabled to see it it behaves the same.

    2nd Test but with Predictable names disabled and this time, the upgrade didn't change the name of the eth0:


      



    Again, never lost ssh or connection on the network.

    For me, and regarding Pis upgrade, having Predictable names disabled via raspi-config PRIOR to the upgrade will prevent the wired interface from changing its name.


    macom

    Does armbian-config exists on your system? Does it have any thing similar to raspi-config to disable the "predictable names"?

    I also see that you have STATIC IP && DNS assigned on the other thread: I'll redo again everything with STATIC IP && DNSes to see what happens.

    • Offizieller Beitrag

    Does armbian-config exists on your system?

    Yes, but I did not find a setting for this. BUT I found this

    Predictable Network Interface names
    Hello, I just installed Armbian on my Tinker board. I use the internal wifi and another wifi usb dongle (static ip). The wifi usb receives a name wlx**** in…
    forum.armbian.com

  • I'm in the office today but will try tomorrow with predictable names disabled.


    I have noticed that when I run raspi-config it asked to define/set the user and defaults to the pi account but I have deleted this account from my omv6 setup a long time ago (as I don't use the pi account/user). Strange that the system still thinks the account exists. The change seems to have 'stuck' after a reboot so I think it will work. I'll post an update tomorrow.

  • I can also confirm the upgrade to omv7 on my rpi4 worked great after disabling predictable network names.


    Omv7 is very stable so far on rpi4. All my prod dockers and services are running fine. I will probably stay on omv7 beta with my omv6 disk/image to go back to if needed.

Jetzt mitmachen!

Sie haben noch kein Benutzerkonto auf unserer Seite? Registrieren Sie sich kostenlos und nehmen Sie an unserer Community teil!